| #include <stdio.h> | |
| #include <thread> | |
| #include <chrono> | |
| #include <iostream> | |
| const int N = 1 << 20; | |
| __global__ void kernel(float *x, int n) | |
| { | |
| int tid = threadIdx.x + blockIdx.x * blockDim.x; |
| #!/usr/bin/env python | |
| """ | |
| Demonstrate how to pass IPC handles to GPU data between processes in Python. | |
| """ | |
| import ctypes | |
| import numpy as np | |
| import multiprocessing as mp | |
| import zmq |
To remove a submodule you need to:
You might want to read this to get an introduction to armel vs armhf.
If the below is too much, you can try Ubuntu-ARMv7-Qemu but note it contains non-free blobs.
First, cross-compile user programs with GCC-ARM toolchain. Then install qemu-arm-static so that you can run ARM executables directly on linux
| ''' | |
| Memory profiling utilities | |
| ''' | |
| import gc | |
| import inspect | |
| import linecache | |
| import os.path | |
| import sys | |
| import time | |
| import threading |
| //----------------------------------------------------------------------------- | |
| // Desc: Dining Philosophers sample. | |
| // | |
| // Demonstrates how to use Monitor object (lock) to protect critical section. | |
| // | |
| // Scenario is such that there are 5 philosophers and 5 tools on each side of a philosopher. | |
| // Each philosopher can only take one tool on the left and one on the right. | |
| // One of the philosophers must wait for a tool to become available because whoever grabs | |
| // a tool will hold it until he eats and puts the tool back on the table. | |
| // |
| // Just before switching jobs: | |
| // Add one of these. | |
| // Preferably into the same commit where you do a large merge. | |
| // | |
| // This started as a tweet with a joke of "C++ pro-tip: #define private public", | |
| // and then it quickly escalated into more and more evil suggestions. | |
| // I've tried to capture interesting suggestions here. | |
| // | |
| // Contributors: @r2d2rigo, @joeldevahl, @msinilo, @_Humus_, | |
| // @YuriyODonnell, @rygorous, @cmuratori, @mike_acton, @grumpygiant, |
| # -*- mode: dockerfile -*- | |
| # Work in progress, some of the manual steps below will be fixed in a subsequent release. | |
| # Dockerfile to build libmxnet.so, and a python wheel for the Jetson TX1 and TX2 | |
| # Builds from Github MXNet master branch | |
| # Once complete copy artifacts from /work/build to target device. | |
| # Install by running 'pip wheel name_of_wheel.whl' and copying the .so to a folder on your LD_LIBRARY_PATH | |
| FROM nvidia/cuda:8.0-cudnn5-devel as cudabuilder | |
| FROM dockcross/linux-arm64 |
| /* https://spectreattack.com/spectre.pdf */ | |
| #include <stdio.h> | |
| #include <stdlib.h> | |
| #include <stdint.h> | |
| #ifdef _MSC_VER | |
| #include <intrin.h> /* for rdtscp and clflush */ | |
| #pragma optimize("gt",on) | |
| #else | |
| #include <x86intrin.h> /* for rdtscp and clflush */ | |
| #endif |
| import numpy as np | |
| cimport numpy as np | |
| cdef extern from "numpy/arrayobject.h": | |
| void PyArray_ENABLEFLAGS(np.ndarray arr, int flags) | |
| object PyArray_SimpleNewFromData(int nd, np.npy_intp* dims, int typenum, void* data) | |
| void _import_array() | |
| cdef np.ndarray[np.int32_t, ndim=1, mode='c'] from_pointer_to_ndarray(np.npy_intp n_elements, | |
| np.int32_t* values): |